Estimation of the Water Footprint of Wood Construction in Chile Using a Streamlined Input–Output-Based Model

Abstract

Wood construction is often proposed to reduce the construction sector’s greenhouse gas emissions due to its carbon sequestration potential. However, forestry significantly impacts natural water flows and increases water use—a critical concern in Chile. This study evaluates the water footprint of wood construction in Chile, considering direct and indirect water consumption under various scenarios. An input–output model was developed to quantify economic interactions, incorporating a new wood-construction sector based on data from a model house. An environmental extension matrix was also created to account for blue water (groundwater and surface water extraction) and green water (rainwater absorbed from soil) consumption. Future scenarios for the residential building sector were defined based on different growth rates for wood-based construction and current construction methods, and the model was resolved using the scenarios as demand vectors. The results indicate that wood construction’s water footprint is 2.38–2.47 times higher than conventional construction methods, with over 64% linked to forestry’s green water demand. By 2050, increased wood construction could raise the sector’s total water footprint by 30.0–31.8%. These findings underscore the need to assess water consumption as a critical sustainability parameter for wood construction and highlight the value of tools like the water footprint to guide decision-making.

1. Introduction

Chile is a country with a low contribution to global greenhouse gas (GHG) emissions, recording a total of 105.552 kt CO₂ equivalent in 2019 [1], which represents only about 0.2% of global emissions. Despite this low contribution, Chile has committed to a National Climate Change Action Plan, with the goal of mitigating and reducing the carbon footprint of its activities [2]. Various economic sectors—such as transportation, industry, agriculture, and energy production—have adopted new strategies to reduce GHG emissions, as outlined in national policies like the Climate Change Framework Law. This law aims to address climate challenges by promoting low-emission development and achieving carbon neutrality by 2050 [3]. Complementing this, Chile’s Long-Term Climate Strategy emphasizes the importance of implementing Sectoral Mitigation Plans and Sectoral Adaptation Plans. These plans outline specific actions and measures to reduce or absorb GHG emissions, meet sectoral emission budgets, and adapt to the impacts of climate change across different economic activities [4]. In the case of the construction sector, one of the alternatives that have been promoted is the use of wood as a building material for various types of structures [5]. The advantages of wood over other materials are diverse, with its role as a carbon sink being a key highlight [6].

In Chile, two key strategies and programs for sustainable construction have been adopted over the last decades. First, the 2013 National Sustainable Construction Strategy, which aims to establish guidelines to promote the incorporation of sustainability criteria in construction. Second, the “Construye 2025” Program, which seeks to transform the way buildings are constructed nationwide, to improve the industry’s productivity across its entire value chain, and to include the life cycle impact of buildings and people’s well-being in their development.

On the other hand, one of the main challenges that Chile faces regarding climate change is related to the development of effective adaptation strategies to minimize its negative effects. In the last decade, the country has experienced a prolonged drought, which has exacerbated water scarcity in various regions. In response to this issue, measures have been implemented to adapt to the water crisis and climate change, aiming to promote the sector’s contribution to the economic, ecological, and socio-cultural development of the country through the comprehensive management and rational use of resources, watersheds, and forest ecosystems. However, forest plantations can affect surface and groundwater flows, influencing the availability of water resources at the watershed level [7]. Moreover, some reports show that forest plantations consume more water than native forests, resulting in lower water yield in the watersheds where they are established [7]. Therefore, it is crucial to evaluate the impact that wood usage in construction can have on wood demand and its indirect effects on water resource availability, particularly given the severe water scarcity the country is currently experiencing.

Despite the potential effect that the development of wood construction may have over wood demand, and therefore the availability and state of water resources, most previous studies focusing on evaluating the environmental sustainability of wood construction do not take this aspect into account. While some research has emphasized the importance of considering environmental impacts of construction in a holistic manner [8], other studies that specifically address the sustainability of the timber industry and the wood construction sector tend to focus on aspects such as GHG emissions, energy use, or forest management [9]. Previous research related to the environmental impacts associated with construction tends to primarily focus on materials such as concrete, steel, and bricks, and the impacts over water resources are generally excluded from the assessment [10].

The connection between wood construction and water resources is an area that requires greater attention. This work aims to address this gap by exploring the potential impacts that increased wood use could have on water resources in Chile. To do this, a streamlined model based on input–output matrices was built and used to test different development scenarios in Chile, providing water footprint results at a national scale for blue water and green water. These results are more relevant from an environmental perspective than the calculations expressed per production unit used in most of the previous studies (e.g., per constructed m²). Furthermore, models based on input–output matrices have the capacity to avoid the truncation error that is prevalent in methods such as process-based Life Cycle Assessment (LCA), which results from cutting-off flows or life-cycle stages during the boundary selection [11].

Water footprint refers to the total amount of freshwater used to produce the goods and services consumed by an individual, community, or country [12,13]. Input–output matrices provide detailed analyses of economic interactions and resource flows between different economic sectors. This connection allows for the quantification and tracking of water consumption throughout the entire supply chain. Input–output methods have been previously used to quantify the water footprint at city and national levels (e.g., [14]), but to the best of our knowledge, it has never been used to assess the water consumption of different development scenarios of the construction sector.

In this study, an environmentally extended input–output model was used to assess water consumption of the residential building construction sector in Chile. The objective was to estimate the total water consumption associated with wood construction and compare it to conventional residential construction under different scenarios of development.

2. Materials and Methods

To meet the study’s objective, a streamlined model was developed using the most recent input–output matrix published by the Central Bank of Chile. This matrix was modified to incorporate a new economic sector representing wood-based construction, and an environmental extension was added to account for the direct consumption of blue and green water in the different economic sectors. Posteriorly, development scenarios for residential construction from 2020 to 2050 were defined, reflecting varying contributions from the conventional construction sector and the newly integrated wood construction sector to the activity of the residential building construction sector. The total water footprint for these scenarios was then calculated using the principles of environmentally extended input–output (EEIO) models.

The Leontief matrix is a widely used tool in economics to analyze the interactions between the sectors of an economy. Equation (1) presents the general expression for EEIO models, which are an extension of the basic Leontief model [15]. These models incorporate an additional matrix that enables the quantification of both direct and indirect emissions or resource consumption tied to a specific level of final demand.

$${\mathrm{e}=\mathrm{B}(\mathrm{I}-\mathrm{A})}^{-1}\mathrm{f}$$

(1)

where

e: Vector representing the total emissions of pollutants or resource consumption associated with the assessed demand (in physical units, such as kg or m³).

B: Environmental extension matrix, representing the direct emissions of pollutants or resource consumption required for producing one unit of monetary output in the corresponding economic sector (in physical units per monetary unit, e.g., kg/CLP).

I: Identity matrix (dimensionless).

A: Direct requirements matrix, indicating the monetary input needed from all economic sectors to produce one unit of output in the corresponding sector (in monetary units).

f: Vector of final demand for the economic sector(s) under assessment (in monetary units).

These types of models have been widely applied to assess emissions and other environmental impacts at national and subnational levels, as well as to evaluate how changes in demand influence the economy’s overall environmental footprint. A notable example in Chile is the study by Avilés-Lucero et al. (2021) [16] which focused on quantifying the country’s carbon footprint, analyzing the contributions of the different productive sectors, and identifying potential reduction strategies. This particular study found that electricity generation is the largest contributor to carbon emissions, followed by the manufacturing industry.

2.1. Modifications Performed on the National Input–Output Matrix

To build the direct requirements matrix (A), the input–output matrix from the Central Bank of Chile was first adjusted. The original matrix consists of 111 × 111 productive sectors, but in order to reduce the matrix to a more manageable, yet detailed number of sectors for which water consumption data were available, 33 broader sectors were defined and their economic input and output were calculated. This reduction was necessary because the environmental extension matrix must have the same number of sectors as the economic matrix in order to solve the model.

The matrix was simplified as follows: first, the economic activities were selected from the glossary defined for the original matrix and then grouped based on their similarities, using the International Standard Industrial Classification (ISIC) codes as a reference. This process resulted in a table with 33 newly aggregated economic sectors. The details regarding the resulting matrix can be found in Table S1 of the Supplementary Materials accompanying this paper.

Since the wood construction sector was not defined in the original input–output matrix, it was necessary to create it. To accomplish this, a detailed materials inventory was obtained from a previously published study that focused on a similar socioeconomic context [17]. This study offers a comparative LCA of a model timber house measuring 62 m² against a concrete-masonry single-family house. The materials were categorized according to their corresponding economic sectors and incorporated into the new A matrix, expressed as monetary requirements based on basic retail selling prices and the estimated value of the model timber house, which corresponds to USD 80,000 [17]. In the case of the monetary inputs necessary from the various service activities of the economy to the newly established wood construction sector, those were assumed to be equivalent to those used by the conventional construction sector.

It is important to mention that given the nature of the input–output matrix used in this study, wood is also considered a part of the materials included in the conventional construction sector of the model. According to data from [18], wood accounted for 16.9% of the residential square meters built in Chile (considering the dominant material used in walls), ranking second after concrete, which represents 53.3%. Brick accounts for 10.4%, while the remaining 19.4% is distributed among materials such as ferrocement panels and combinations arising from the previously mentioned materials (wood-brick, ferrocement-wood panels, concrete-wood, and others).

2.2. Construction of the Matrix of Direct Water Consumption

The second activity involved determining the water consumption of the economic sectors in matrix A, utilizing various information sources. For most of the primary and secondary sectors, one to three representative production processes were identified, with water consumption data primarily drawn from the Ecoinvent 3.6 database. These data were then expressed in volume per monetary unit based on the basic retail selling prices of the products generated by the selected processes, averaged to represent the overall consumption of the sector. Detailed information regarding the processes and their specific water consumption can be found in Table S2 of the Supplementary Materials accompanying this paper.

For tertiary sectors such as public services, tourism, and others, it was assumed that the most significant fraction of water consumption in these sectors originates from persons, and therefore, water consumption was estimated based on the number of employees in each sector and specific consumption rates derived from the literature [19]. Employee data for each sector was obtained from company statistics published by the Internal Revenue Service (SII) [20], and water consumption was expressed in monetary units by dividing the total estimated consumption by the Gross Domestic Product (GDP) of each specific sector extracted from the original input–output matrix. Details regarding the estimation of water consumption for the tertiary sector of the economy can be found in Table S3 of the Supplementary Materials accompanying this paper.

In order to calculate water use during agricultural and forestry activity, the blue and green water footprint (WFblue and WFgreen, respectively) proposed by Hoekstra et al. (2011) [21] was used. For this, the following equations were used:

$$\mathrm{E}\mathrm{T}\mathrm{b}\mathrm{l}\mathrm{u}\mathrm{e}=\mathrm{E}\mathrm{T}\mathrm{c}-\mathrm{P}\mathrm{e}\mathrm{f}\mathrm{f}$$

(2)

$$\mathrm{E}\mathrm{T}\mathrm{c}={\mathrm{E}\mathrm{T}}_0\times \mathrm{K}\mathrm{c}$$

(3)

$$\mathrm{E}\mathrm{T}\mathrm{g}\mathrm{r}\mathrm{e}\mathrm{e}\mathrm{n}=\mathrm{P}\mathrm{e}\mathrm{f}\mathrm{f}$$

(4)

$$\mathrm{W}\mathrm{F}\mathrm{b}\mathrm{l}\mathrm{u}\mathrm{e}=\mathrm{E}\mathrm{T}\mathrm{b}\mathrm{l}\mathrm{u}\mathrm{e}\times {\mathrm{Y}}^{-1}$$

(5)

$$\mathrm{W}\mathrm{F}\mathrm{g}\mathrm{r}\mathrm{e}\mathrm{e}\mathrm{n}=\mathrm{E}\mathrm{T}\mathrm{g}\mathrm{r}\mathrm{e}\mathrm{e}\mathrm{n}\times {\mathrm{Y}}^{-1}$$

(6)

ETblue and ETgreen represent evapotranspiration (m³) from irrigation water and rainwater, respectively. ETc is the crop evapotranspiration (m³). ET₀ represents the potential evapotranspiration (m³), which was sourced from INIA [22] and whose calculation is based on the Penman–Monteith method of the Food and Agriculture Organization (FAO). Kc (dimensionless) is the crop coefficient, which was calculated on a monthly basis as established by [23]. Peff represents the effective precipitation, which was determined using the simplified USDA method [24]. Y is the yield of each crop (kg per hectare), which was obtained from Chilean government organizations [25].

In addition to the calculation based on WFN recommendations, a “Net Green” (NG) approach was employed to estimate the green water footprint [26]. For this approach, Equations (2)–(6) were also applied. The key difference lies in the calculation of ETgreen, which specifically represents the precipitation utilized by crops that would not have been required by the potential natural biomass of the site. In other words, the net green water appropriated by agroforestry activities is determined by the difference between ETc and ETo, provided that effective precipitation (Peff) meets the crop’s water requirements. However, if Peff falls short, ETgreen is calculated as the difference between Peff and ETo, since an additional input of water as irrigation (blue water) would then be needed for optimal crop growth and development. These conditions are defined as follows (Equations (7) and (8)):

$$\mathrm{If} \mathrm{P}\mathrm{e}\mathrm{f}\mathrm{f}>\mathrm{E}\mathrm{T}\mathrm{c}>\mathrm{E}\mathrm{T}\mathrm{o}; \mathrm{E}\mathrm{T}\mathrm{g}\mathrm{r}\mathrm{e}\mathrm{e}\mathrm{n}=\mathrm{E}\mathrm{T}\mathrm{c}-\mathrm{E}\mathrm{T}\mathrm{o}$$

(7)

$$\mathrm{If} \mathrm{E}\mathrm{T}\mathrm{c}>\mathrm{P}\mathrm{e}\mathrm{f}\mathrm{f}>\mathrm{E}\mathrm{T}\mathrm{o}; \mathrm{E}\mathrm{T}\mathrm{g}\mathrm{r}\mathrm{e}\mathrm{e}\mathrm{n}=\mathrm{P}\mathrm{e}\mathrm{f}\mathrm{f}-\mathrm{E}\mathrm{T}\mathrm{o}$$

(8)

Details regarding the results of water consumption estimation for agriculture and forestry can be found in Table S4 of the Supplementary Materials accompanying this paper.

2.3. Scenario Definition

The third activity involves quantifying the water footprint of timber construction under different scenarios. This analysis assesses the direct and indirect water consumption associated with various development strategies for timber construction, allowing for comparisons with conventional residential construction and the identification of critical points in the supply chain. For scenario definition, a time horizon between 2020 and 2050 was established, starting with an initial condition in which timber construction accounts for 0% in 2020. For 2021, a market share of 1.17% (S0) for timber construction was assumed, based on annual statistics provided by the Chilean Chamber of Construction (CChC) through the Monthly Construction Activity Index (IMACON). From 2021 onwards, the entire construction sector is assumed to grow at the same rate (1.17%) in the baseline scenario (S0), but higher annual growth rates for timber construction are assumed for other scenarios: 2% in the first scenario (S1), 4% in the second scenario (S2), and 6% in the third scenario (S3). The different scenarios were incorporated in the model as final demand vectors (f), and the model was resolved in time steps of five years.

Table 1. Final demand of the residential building sector considering conventional construction and wood-based construction for each of the evaluated scenarios. All values are in billions of Chilean pesos (10⁹ CLP).

Scenarios		2020	2025	2030	2035	2040	2045	2050
S0	CC	5904.0	6256.7	6631.4	7028.5	7449.4	7895.5	8368.3
	WC	0.0	74.1	78.5	83.2	88.2	93.5	99.1
S1	CC	-	6200.0	6565.5	6952.2	7361.5	7794.6	8252.8
	WC	-	130.8	144.4	159.5	176.1	194.4	214.6
S2	CC	-	6048.0	6379.1	6693.1	7028.3	7369.4	7713.6
	WC	-	282.8	330.8	418.6	509.3	619.6	753.8
S3	CC	-	5873.0	6097.3	6292.0	6440.6	6520.9	6502.8
	WC	-	457.7	612.6	819.8	1097.0	1468.1	1964.6

CC: conventional construction; WC: wood construction.

summarizes final demand vectors for every modeled scenario, in billions of pesos (10⁹ CLP).

2.4. Contribution Analysis

A contribution analysis was performed to identify the most significant sectors in terms of their contribution to the total, green, and blue water footprint. In order to do so, the environmental extension matrix B was diagonalized, placing the values of direct blue and green water consumption on the diagonal of a new matrix, with all other values set to 0. By solving the model with this new matrix, the water footprint for each sector is obtained. In order to identify the hotspots of blue water footprint in the model supply chains, the Structural Path Analysis (SPA) method was applied to the results. SPA is a method used to trace how economic activities are interconnected within an input–output framework. It decomposes the input–output model in individual supply chains or paths, and quantifies its contribution to the total economic output of the model or the associated footprints [27]. In the case of EEIO models, SPA can be used to identify the sectors that show the highest environmental impact or footprint in an economic system, and to identify how far upstream from the target sector these impacts are occurring. spaJS was used for this analysis, which is a freely available and online tool developed in JavaScript by researchers at The University of Sydney [28]. SPA was performed for the blue water footprint of both the conventional and wood-based construction, using a pruning threshold of 1 × 10⁻⁸ L/CLP that results in a maximum footprint coverage of 95.74% and 99.95%, respectively.

3. Results and Discussion

3.1. Comparison of the Water Footprint of Conventional Construction and Wood Construction

The results show that the water footprint associated with building a reference timber house of 62 m² is 3212.7 m³ according to the WFN method, while the NG method records a footprint of 2309.9 m³. In contrast, the calculated water footprint of a conventional house is 1302.3 m³ using the WFN method and 968.7 m³ with the NG method. These values suggest that timber construction demands 2.38–2.47 times more water compared to conventional construction, depending on the calculation method. Our findings are consistent with previous studies, which reports water footprint values between 12.64 and 54.10 m³ per constructed m² [29,30]. In terms of the contribution of the two calculated water footprint components, the green water footprint for the timber house accounts for 81% of the total according to the WFN method and 73% according to the NG method. In contrast, the conventional house shows a green water footprint of 74% based on WFN and 64% according to the NG approach.

The lower environmental footprint of conventional construction compared to timber construction contradicts the findings of previous studies that emphasize other environmental indicators. Among these, the most commonly reported indicator in the literature is GHG emissions and/or Global Warming Potential (GWP) [31]. This preference arises from wood’s ability to act as a carbon reservoir and the lower energy demands associated with its processing into building materials. A 2022 review by Duan et al. [32] revealed that the average embodied GHG emissions of reinforced concrete buildings are 42.68% higher than those of mass timber alternatives. Similarly, Kumar et al. (2024) [33] reported that mass timber buildings exhibit 81–94% lower GWP than concrete buildings and 76–91% lower GWP than steel buildings. Supporting these findings, a study conducted in Chile demonstrated that wood-based construction produces 0.12 tons of CO₂eq per square meter, while the construction of a house using reinforced concrete generates 0.21 tons of CO₂ eq per square meter [34]. The fact that the water footprint results contradict other environmental indicators highlights the need to consider this metric when evaluating the environmental sustainability of wood construction, in order to avoid or minimize the potential environmental trade-offs that could arise from promoting this strategy on a large scale.

3.2. Influence of Timber Construction Development on the Water Footprint of the Residential Building Sector

Figure 1 presents the water footprint results for the various scenarios, based on the two calculation methods. The graph on the left (Figure 1a) shows a continuous upward trend in the total water footprint across all scenarios over time, indicating a progressive increase in water consumption. As expected, S3 exhibits the highest projected water footprint, reaching approximately 150 million m³ per year by 2050, while scenario S0 shows the lowest footprint under all conditions. Similarly, the graph on the right in Figure 1 (Figure 1b) displays a comparable pattern of gradual growth, although at a more moderate rate than the graph on the left. Once again, scenario S3 records the highest water footprint, whereas scenario S0 reflects the lowest, reinforcing the conclusion that increased adoption of timber construction significantly impacts water usage over time. By 2050, the analysis of the scenarios suggests that S3 could lead to an increased total water footprint of the sector by 30.0–31.8% compared to S0.

Table 2. Contribution of green water footprint to the total water footprint of the construction sector for all the assessed scenarios.

Scenarios	2020		2025		2030		2035		2040		2045		2050
	WFN	NG	WFN	NG	WFN	NG	WFN	NG	WFN	NG	WFN	NG	WFN	NG
S0	74.4%	65.4%	73.7%	64.7%	73.7%	64.7%	73.7%	64.7%	73.7%	64.7%	73.7%	64.7%	73.7%	64.7%
S1	-	-	73.9%	64.8%	73.9%	64.9%	73.9%	64.9%	73.9%	64.9%	74.0%	64.9%	74.0%	64.9%
S2	-	-	74.3%	65.3%	74.4%	65.4%	74.5%	65.6%	74.6%	65.7%	74.8%	65.9%	75.0%	66.1%
S3	-	-	74.7%	65.8%	75.0%	66.1%	75.3%	66.5%	75.7%	67.0%	76.1%	67.5%	76.7%	68.2%

WFN: according to the water footprint network method; NG: according to the net green method.

presents the contribution of the green water footprint across the previously outlined scenarios. In the baseline scenario, the contribution of the green water footprint ranges from 73.7% to 74.4% when using the WFN calculation method, remaining constant from 2025 onward. In contrast, when applying the NG method, the contribution of the green water footprint is between 64.7% and 65.4%. For the other scenarios, the contribution of the green water footprint varies from 73.9% to 76.7% according to the WFN method and from 64.8% to 68.2% using the NG method. As expected, the WFN method results in higher contributions from the green water footprint because it allocates the total evaporative flow of the forest crop to productive activities. In comparison, the NG method only assigns the evapotranspiration associated with the modification between the crop and a reference scenario. Overall, scenarios with a higher proportion of wood construction yield a greater contribution to the green water footprint, with the maximum value achieved in scenario S3 by the year 2050, regardless of the calculation method used. This is linked to the increased demand for timber in this scenario.

3.3. Analysis of the Contributions to the Water Footprint

The results of the contribution analysis reveal that, as expected, the green water footprint of wood construction predominantly originates from the wood and forestry sector, which account for 99.7% of the total. In contrast, the agricultural sector contributes only 0.3%. For conventional construction, the wood and forestry sector similarly contributes 99.6%, while the agricultural sector accounts for 0.4%. Figure 2 illustrates the contribution analysis for the blue water footprint in wood construction. The data show that for wood construction the largest contribution comes from the sawmilling and wood products sector, accounting for 47%, followed by fishing and aquaculture, which contribute 27%. Agriculture contributes 9%, while the paper and cardboard sector accounts for 5%, making these the primary economic sectors responsible for the blue water footprint. In the case of conventional construction, the results indicate that the fishing and aquaculture sector represents the largest contribution at 38%, followed by sawmilling and wood products at 30%, agriculture at 9%, and the paper and cardboard sector at 7%.

Table 3. Percent contribution (%) of economic sectors and supply chain levels to the blue water footprint of residential building construction in timber. Only the results of the 10 economic sectors with the highest contribution are shown.

Sector/Level	1	2	3	4	5	6	7	8	9	TOTAL
Sawmills and wood products	39.69	6.00	1.03	0.03	0.02	0	0	0	0	46.77
Fishing and aquaculture	0.03	7.06	7.6	4.32	4.63	1.70	0.58	0.16	0.04	26.12
Agriculture	0.01	0.95	3.48	0.91	2.51	0.24	0.06	0.01	0	8.17
Paper and cardboard	0.84	2.21	1.25	0.06	0.39	0.01	0	0	0	4.76
Electricity	0.96	0.94	0.38	0.01	0.09	0	0	0	0	2.38
Paintings	1.89	0.17	0.05	0	0.01	0	0	0	0	2.12
Commerce	1.05	0.72	0.24	0.01	0.06	0	0	0	0	2.08
Pharmaceutical products	0.28	0.96	0.13	0	0.02	0	0	0	0	1.39
Metals	0.18	0.72	0.16	0	0.03	0	0	0	0	1.09
Agroindustry	0.14	0.14	0.08	0.03	0.08	0.01	0	0	0	0.48
TOTAL	45.07	19.87	14.4	5.37	7.841	1.96	0.64	0.17	0.04	95.36

and

Table 4. Percent contribution (%) of economic sectors and supply chain levels to the blue water footprint of conventional residential building construction. Only the results of the 10 economic sectors with the highest contribution are shown.

Sector/Level	1	2	3	4	5	6	7	8	9	TOTAL
Fishing and aquaculture	0.05	12.6	10.42	4.26	6.78	1.68	0.51	0.12	0.02	36.44
Sawmills and wood products	21.76	6.11	1.32	0.03	0.28	0	0	0	0	29.5
Agriculture	0.02	1.69	2.97	0.8	2.1	0.21	0.04	0	0	7.83
Paper and cardboard	1.5	3.15	1.67	0.07	0.5	0	0	0	0	6.89
Electricity	1.7	1.08	0.54	0.01	0.13	0	0	0	0	3.46
Metals	2.03	1.14	0.25	0	0.04	0	0	0	0	3.46
Commerce	1.87	0.93	0.33	0.01	0.08	0	0	0	0	3.22
Pharmaceutical products	0.5	0.75	0.17	0	0.03	0	0	0	0	1.45
Agroindustry	0.26	0.18	0.12	0.03	0.07	0.01	0	0	0	0.67
Professional services	0.22	0.1	0.04	0	0.01	0	0	0	0	0.37
TOTAL	29.91	27.73	17.83	5.21	10.02	1.9	0.55	0.12	0.02	93.29

present the blue water percentage contributions of the economic sectors at different supply chain levels for wood based and conventional construction, respectively. This evaluation considers the ten economic sectors with the highest contributions across nine levels of the supply chain. Table 3 reveals that for wood construction the sawmill and wood products sector contributes the highest percentage (39.69%) at the first level of the supply chain, reflecting the water requirements necessary for transforming wood into the various construction elements used by the industry (such as panels, beams, and flooring). On the other hand, the results indicate that most of the contribution from the fishing and aquaculture sector (23.61%) is distributed across levels 2 to 5 of the supply chain, with levels 3 and 2 standing out with contributions of 7.60% and 7.06%, respectively. Similarly, Table 4 shows that for conventional construction, the contribution from the sawmill and wood products sector at the first level of the supply chain amounts to 21.76%, with an additional 7.74% spread across levels 2 to 5. In the case of the fishing and aquaculture sector, the distribution mirrors that of wood-based construction, with 34.06% allocated across levels 2 to 5, and the highest contributions coming from levels 2 and 3.

The high contribution of the fishing and aquaculture sector is primarily associated with two factors. First, the relevance of agro-industry in the structure of Chile’s economy and its extensive interactions with other economic sectors. Agro-industry processes raw materials from primary activities such as agriculture, livestock, fishing, and aquaculture, creating notable interconnections within the input–output matrix, especially with the fishing and aquaculture sector. This results in a significant indirect contribution to the water footprint of many other economic sectors. Second, the sector’s specific water consumption, as estimated in the B matrix developed for this study, is the highest among all sectors. For this study, it was assumed that the water consumption of the fishing and aquaculture sector could be represented by the water usage of salmon aquaculture, given the economic significance of this activity in Chile. Although the assumed water consumption for salmon aquaculture was set at a conservative value of 14.82 m³/kg (sourced from the Ecoinvent 3.6 database), the consumption per monetary unit amounted to 1.56 L/CLP. Water consumption in salmon aquaculture is primarily associated with the fry and smolt stages, which take place in freshwater environments such as lakes or river-based hatcheries. This indicates a significantly higher value compared to other sectors. For instance, the agriculture sector reported a value of 0.3442 L/CLP, which, while representing the second-highest blue water consumption across the entire B matrix (Table S5 of the Supplementary Materials), is still 4.5 times lower than the value for fishing and aquaculture.

3.4. Limitations of This Study and Future Challenges

It is important to acknowledge several limitations of this study considering its objectives. First, the reduction in economic sectors in the direct requirements matrix led to a lower level of detail in identifying critical points. While this simplification was necessary as an initial step to facilitate the construction of the environmental extension matrix, future studies should aim to use matrices with greater sectoral detail. Furthermore, the inclusion of the wood construction sector in matrix A was accomplished by creating a new sector, without disaggregating the residential construction sector into its components: one that primarily uses wood as building material and another that relies on materials such as concrete. Although this approach is standard practice when working with input–output matrices, it was not applied here due to insufficient reliable data for such disaggregation. Additionally, the model does not explicitly account for key elements typically included in input–output analyses, such as export and import flows or capital formation, which should be incorporated into future studies. Finally, the level of detail used for estimating water consumption in the B matrix was not the same across sectors, which could result in some values being overestimated or underestimated. This last issue could be mitigated by conducting sensitivity analyses in future studies, in order to evaluate these and other aspects of the model. Sensitivity analyses would also help address the effects of price volatility and the uncertainty inherent in working with this type of data.

4. Conclusions

The results of this study suggest that wood-based construction in Chile exhibits a higher water footprint compared to residential construction based on currently used materials. Specifically, it has been identified that the water footprint associated with wood construction can be between 2.38 and 2.47 times higher than conventional construction, depending on the calculation method. Furthermore, the analysis of projected scenarios suggests that promoting the use of wood as a construction material in Chile could increase the sector’s total water footprint by 30.0–31.8% by 2050.

One of the most important findings of this study is the critical influence of the timber products supply chain on total water consumption of construction, considering both the green and blue water footprints. This study demonstrated that, regardless of the type of construction, the scenario evaluated, or the method used to calculate the green footprint, its contribution is consistently greater than that of the blue water footprint, accounting for between 64.7% and 76.6% in the scenarios analyzed. This underscores the importance of implementing water-use optimization policies in the forestry sector as a measure to mitigate the impact of construction on water resources.

Moreover, for both conventional and wood-based construction the sectors with the highest contribution to the blue water footprint were sawmills and wood products, as well as fishing and aquaculture. In the case of the sawmill and wood products sector, the results of the Structural Path Analysis indicates that the processing of materials directly used in construction had the largest contribution to water footprint, whereas the contribution of the fishing and aquaculture sectors was primarily distributed across supply chain levels 2 and 5. This finding suggests that such consumption arises mainly from indirect interactions, particularly through the agro-industrial sector.

In conclusion, our research reinforces the need to consider water consumption as a critical factor in promoting more sustainable construction strategies. For wood-based construction, its sustainability in terms of the water footprint depends significantly on the adoption of strategies that enhance water-use efficiency throughout the entire supply chain.